Systemic immunization of BALB/c mice with human cartilage proteoglycan (PG) aggrecan induces progressive polyarthritis. This murine model (PG-induced arthritis;PGIA) shares similarities with rheumatoid arthritis (RA) as indicated by clinical assessments, laboratory tests, x-ray and histopathology of diarthrodial joints. Similar to RA, PGIA is a T cell-dependent and antibody/B cell-mediated autoimmune disease. Association of age and gender with disease development and severity, and recessive inheritance of disease susceptibility, dictated by both the major histocompatibility complex (MHC)- and non-MHC-associated genes are additional characteristics that make this model an ideal tool for genetic studies. While the MHC is a critical predictive factor both in RA and PGIA, the MHC alone is insufficient for the induction of autoimmune disease. Although large-scale genome-wide association (GVA) studies repeatedly showed linkage between certain chromosome regions and RA, not a single gene has been identified in the heterogeneous human population as a disease-inducing gene. During the past 10 years, we tested over 3,200 F2 hybrid (including 1,292 arthritic) mice of six different genetic intercrosses. A total of 29 genomic loci (Pgia1-Pgia29 ) controlling PGIA were identified. Many of these quantitative trait loci (QTLs) correlated with those identified in collagen-induced arthritis (CIA) in mice (mCia) and rats, colocalized with QTLs of other autoimmune models, and corresponded with human genomic regions identified in GWA studies of RA. Our hypothesis is that if a genomic locus is shared among different autoimmune or arthritis models, it is likely involved in similar immune regulatory pathways, which also operate in RA and perhaps in autoimmune diseases in general. We intentionally excluded the MHC effect using two murine strains carrying an identical H2d MHC allele, and we have generated five congenic strains containing DBA/2 (PGIA-resistant strain) alleles in chromosomes 3, 7, 8, 15 and 19 in BALB/c (PGIA-susceptible strain) background. The major QTLs were selected to correspond with syntenic regions of the human genome, where several RA-associated QTLs have been identified. After initial testing of these congenic and then subcongenic mouse strains, we selected Pgia26 on chromosome 3 (chr3) for subsequent studies to narrow the chromosome interval to a manageable size, and identify disease-suppressive (within the Pgia26a locus) and disease-promoting (in Pgia26b locus) genes. We selected Pgia26, because it appears to possess one of the most complex loci (after the MHC) regulating arthritis susceptibility and severity in PGIA, and the corresponding (syntenic) region on human chr1 (Chr1:95.2-151.7 Mbp region, including PTPN22) shows the most prominent linkage with RA, after the MHC. In the genomic research history of experimental arthritis, we have identified probably the smallest size of arthritis-associated genomic regions, and now we are in the process of selecting interval-specific congenic (IVSC) strains to reduce further the size of these critical loci. We propose two specific aims to select IVSC strains in both disease-suppressing Pgia26a (chr3:90.4/92.7-96.4/99.9 Mbp) and arthritis-promoting Pgia26b (chr3:108.1/109.2-115.8/121.1 Mbp position) to reduce the chromosome region to a manageable size (Aims 1A and 2A), and then prepare detailed physical maps and introduce positional candidate gene cloning (Aims 1B and 2B). Selected genes will be tested for their expression in arthritic joints and joint-draining lymph nodes, sequenced, and their in vivo (arthritis-affecting) function will be tested in either knockout or IVSC mice, carrying structurally and/or functionally altered gene(s) and altered disease phenotype. PUBLIC HEALTH RELEVANCE: Rheumatoid arthritis (RA) is a complex joint disease affecting approximately 1 % of the human population. Large scale human familial genome-wide association studies have localized chromosome regions associated with RA, but no individual "causative" genes had been identified in the heterogeneous human population within these chromosome regions. We have developed a mouse model of RA, which shows many similarities to the human disease. We narrowed the genomic (chromosome) regions of a number of disease-associated regions, and selected a segment of mouse chromosome 3, which is highly associated with disease phenotype in our arthritis model. This chromosome region (Pgia26) contains both arthritis-promoting and -suppressing genes, and corresponds to a region of human chromosome 1, that shows the highest association with RA after the major histocompatibility complex (MHC). We propose to reduce the size of these regions in the mouse genome, which cannot be done in human population, and select candidate genes that may be involved in arthritis susceptibility and disease severity.